Published online by Cambridge University Press: 17 December 2009
Based on a 1900 census sample of 34,166 post-reproductive females (≥45 years), the birth month effect was put to the test, for both lifetime fertility as well as child survival, controlling for maternal birth cohort (1826–1835, 1836–1845, 1846–1855), Duncan's SEI, urbanity, nativity, literacy and marital duration. Testing for potential cohort effects did not indicate a temporal trend in fertility by maternal birth month (seasonal Mann-Kendall test, p=0.578), while a minute increase in offspring survival was detected (p<0.001, Sen's estimator of slope=0.02, 95% CI=0.02 to 0.03). Further analyses of the maternal birth month effect on child survival were therefore seized. For lifetime fertility, ANOVA results indicated that maternal birth month was a major predictor for total offspring count (F11, 33606=1809.0, p<0.001), accounting for 37.2% of the total variability. In addition to main effects, a statistically significant interaction effect was observed (F538, 33606=2.2, p<0.001), with a corresponding effect size of η2=0.40. Planned contrasts revealed that birth-month-specific differences in fertility achieved statistical significance (F11, 31798=1712.9, p<0.001), while post-hoc multiple comparisons for literacy and nativity displayed an inverse relationship with fertility, which meets demographic expectations. Controlling for all factors of interest, models of cohort-specific offspring counts (independent ANOVAs for 1826–1835: F157, 3467=26.3, p<0.001; 1836–1845: F182, 10299=75.5, p<0.001; 1846–1855: F199, 19859=137.9, p<0.001) indicated that women born in the first half of the year (particularly, January, February, April and May) achieved above-average parity, while those born in the latter half (namely, July, October, November and December) displayed markedly lower fertility averages. These monthly disparities are in line with previous observations and appear to be linked to seasonal optimal ripening of the oocyte or seasonal preovulatory over-ripeness ovopathy (Jongbloet, 1992).